Daggerboard Design- a different perspective

[farjoe]

Hi,

I own a 7.5m cat which has about 30 sq m of upwind sail area.It has 2 daggerboards each 1300mm deep and 500 mm span. The daggerbox has a cross sectional area of 500mm by 38mm. This has resulted in a daggerboard profile which approximates a NACA0007.

Previous correspondence suggests that this is not an ideal shape for upwind work and a NACA0012 would be more appropriate.

I would like to improve the boat's pointing characteristics and add a small bit of weather helm to the boat. The boat has a small amount of lee helm.

Would I be able to achieve my targets if I build NACA0012 foils to fit the same boxes? The span would then be about 320mm. By fitting the new boards all the way forward in the boxes the centre of lateral resistance should move forward by about 90mm.

Are these ideas flawed?

Would I have to build the boards deeper?

Your thoughts on this proposal are most welcome.

regards

joe

[water addict]

The NACA0012 shape may not be any better. By span, I think you mean chord length. Span is the vertical length, chord is the distance from leading to trailing edge. If you want more lift, you do not want to decrease the blade area. If you use 0012 section with the same span you will have less blade area, thus reducing lift - probably not good.
I would not recommend a 0012 section for a cat centerboard. You are probably going high speed at relatively small angles of attack if the boat is sailed right. 0012 would be better for a rudder section where you need some forgiveness from stall.
You say you have lee helm- strange. Are you sailing upwind with both boards down? You should be until you're going very fast. If still lee helm, try tensioning the mainsheet some. I think it's strange to have lee helm from a cat, because figuring CE and CLR is pretty straightforward, since heel should be small- you won't get the big moment from the sailplan being to leeward and the keel to windward like a monohull. Is this a standard production cat? If so, talk to others who are experts at racing them to find out their tips before going to the effort of making new blades.

Good luck! (I'm a naval arch. for the US navy)

Originally Posted by farjoe

Hi,

I own a 7.5m cat which has about 30 sq m of upwind sail area.It has 2 daggerboards each 1300mm deep and 500 mm span. The daggerbox has a cross sectional area of 500mm by 38mm. This has resulted in a daggerboard profile which approximates a NACA0007.

Previous correspondence suggests that this is not an ideal shape for upwind work and a NACA0012 would be more appropriate.

I would like to improve the boat's pointing characteristics and add a small bit of weather helm to the boat. The boat has a small amount of lee helm.

Would I be able to achieve my targets if I build NACA0012 foils to fit the same boxes? The span would then be about 320mm. By fitting the new boards all the way forward in the boxes the centre of lateral resistance should move forward by about 90mm.

Are these ideas flawed?

Would I have to build the boards deeper?

Your thoughts on this proposal are most welcome.

regards

joe

[sorenfdk]

Board area is not everything - the aspect ratio (span squared/board area) is just as important!
If you use a 0012 section and shorten the chord to 320 mm, you'll need to increase the span to appr. 1600 mm to achieve the same lift.
The smaller wetted surface means lower drag - somewhere in the region of 20% lower.

[Tim B]

quite true, soren... but one thing springs to mind... two hulls, two daggerboards, assymetric boards using cambered foils perhaps? there are plenty out there to choose from. just raise the windward board when you're sailing, bit more work I know, but it is the most efficient way.

Good luck,

Tim B.

[pmusu]

I don't know about multihulls, however in dinghy's when sailing downwind, for better sailing you should lift the daggerboard.

[sorenfdk]

Originally Posted by Tim B

quite true, soren... but one thing springs to mind... two hulls, two daggerboards, assymetric boards using cambered foils perhaps? there are plenty out there to choose from. just raise the windward board when you're sailing, bit more work I know, but it is the most efficient way.

Good luck,

Tim B.

Damn! Why didn't I think of that?

[farjoe]

hi,

To answer some of the above.....

As was obvious to everyone I should have written "chord" rather than "span".

I attach a comparison performed using the program "Profili" of the 2 profiles. Unless I am mistaken (quite possible) the improvement over the 0007 seems substantial.

Soren, Is this the reason why you suggested a span of 1600mm only for an equivalent lift?

I chose the 0012 as an improvement because I wanted to stay within the 38mm max thickness. This is because I do not want to redo the case. Going for a higher NACA profile would mean an even shorter chord and that seems risky.

How much of an improvement can one reasonably expect when building boards in practice? Unfortunately automated profiling is out.

Referring to comments about the lee helm, it would be better if I describe it as "lee to neutral helm". In this condition a wave hitting the bows in the wrong direction easily pushes the boat off the wind which may be dangerous for singlehanding in a fresh breeze.

I am hoping that some weatherhelm ought to counteract this most times.

regards

joe

[sorenfdk]

I use Profili, too. It's a great piece of software, but remember that it only does 2-D calculations.
For 3-D calculations, I use SailingAerodynamics by Hanley Innovations. I used it to calculate the lift and drag of your current board. Then I changed the section and chord and calculated the span to achieve the same lift. In the calculations, I made some assumptions regarding speed (10 knots) and leeway angle (4 degrees), but that shouldn't have any significant influence.
So yes - the 1600 mm span gives the same lift. Perhaps you should choose 1620 mm just to be sure...

[Sean Herron]

Hello...

I am just now mucking about with the idea of a shallow full keeled monohull with vertically lifted but asymetrical foils to pull the boat upwind...

I would actually like to try same on a cat - I think that Kurt Hughes has done so on his larger racers - also link below may tell you more of what I am up to...

I would be afraid of some hobby horsing with an asymetrical foil on a cat flying a hull...

Anyway here is the link - http://www.selectyachts.co.uk/redfoxhome.asp?ArtID=184 ...


SH

[farjoe]

Originally Posted by sorenfdk

I use Profili, too. It's a great piece of software, but remember that it only does 2-D calculations.
For 3-D calculations, I use SailingAerodynamics by Hanley Innovations. I used it to calculate the lift and drag of your current board. Then I changed the section and chord and calculated the span to achieve the same lift. In the calculations, I made some assumptions regarding speed (10 knots) and leeway angle (4 degrees), but that shouldn't have any significant influence.
So yes - the 1600 mm span gives the same lift. Perhaps you should choose 1620 mm just to be sure...

Soren,

Sailearo seems like a good software but unfortunately expensive for an amateur like me. I guess I have to make do with Profili. Incidentally have you ever compared the results between the 2? Would the inaccuracies introduced when building an actual foil override any theoretical advantages anyway?

I have rerun Profili using 4 degrees and 10 knots( Re = 1600000 for NACA0012 and Re = 2600000 for NACA0007) as you suggest and my conclusions seem to be somewhat different to those arrived at previously.

From this starting point I get the following numbers:

NACA0007 CL=0.457 CD=0.0066
NACA0012 CL=0.433 CD=0.0066

i.e. the NACA0007 seems to be slightly better in these conditions!!!!

In fact 4 degrees angle of attach seems to be the crossover point. Below this figure NACA0007 seems to be better. Above, NACA0012 appears better.

Not being a Mechanical Engineering I wonder if I am taking the wrong assumptions!!


regards

joe

[sorenfdk]

Originally Posted by farjoe

Soren,

Sailearo seems like a good software but unfortunately expensive for an amateur like me. I guess I have to make do with Profili. Incidentally have you ever compared the results between the 2? Would the inaccuracies introduced when building an actual foil override any theoretical advantages anyway?

I have rerun Profili using 4 degrees and 10 knots( Re = 1600000 for NACA0012 and Re = 2600000 for NACA0007) as you suggest and my conclusions seem to be somewhat different to those arrived at previously.

From this starting point I get the following numbers:

NACA0007 CL=0.457 CD=0.0066
NACA0012 CL=0.433 CD=0.0066

i.e. the NACA0007 seems to be slightly better in these conditions!!!!

In fact 4 degrees angle of attach seems to be the crossover point. Below this figure NACA0007 seems to be better. Above, NACA0012 appears better.

Not being a Mechanical Engineering I wonder if I am taking the wrong assumptions!!


regards

joe

As I said before, Profili only does 2-D calculations, while SailAero does 3-D, so you can't really compare.
You can use Profili or any other 2-D software to make comparisons between different sections, but do not use it to make comparisons between boards, as the lift and drag coefficients for 2-D sections are not equal to the 3-D coefficients!
The lift and drag of boards depend on the area (A) and aspect ratio (AR):

CL = 0.11*alpha/(1+2/AR)

CDi = CL**2/(pi*AR)

where
0.11 is the theoretical value of the 2-D lift coefficient per degree of the 2-D section in a frictionless fluid. You might want to use 0.10 instead to compensate for viscosity effects.
Alpha is the angle of attack.
AR is the aspect ratio (= span**2/A).
CDi is the induced drag coefficient.

Of course, real-life inaccuracies such as surface roughness, bumps etc. do play a role. It is possible to calculate the maximum "permitted" roughness and bump height, but I don't remember how. Just make the surface as smooth as you can, and you'll be OK!

[sorenfdk]

SailAero has a built-in 2-D calculator. I don't know what method or formulas it uses, so I can't say anything about how accurate it is. Profili, on the other hand, is built on Xfoil, which has proven itself to be very accurate.
I've done a quick comparison using NACA 0012 at RE = 1000000, and there are differences between the two. Profili/Xfoil predicts higher CL and CD. Max CL/CD is at the same angle of attack, but Profili/Xfoil predicts a higher CL/CD. The CM vs. AoA curves are very different - at an AoA of 5 degrees, Profili/Xfoil predicts CM = 0.005, SailAero says CM = -0.005. At AoA = 14 degrees, the values are CM = 0.028 and -0.12, respectively.
I think I rely more on Profili, but I can't really tell what, if any, impact the - perhaps wrong - results from SailAero's 2-D calculator has on the 3-D results. According to the formulas above, it shouldn't be much.
BTW: In the comparison I used Ncrit = 9 in Profili. This should give values that are comparable to what you would measure in an average wind tunnel. But keels and rudders don't operate in tunnels, so here one should use a lower value, say Ncrit = 3.

[farjoe]

Originally Posted by sorenfdk

The lift and drag of boards depend on the area (A) and aspect ratio (AR):

CL = 0.11*alpha/(1+2/AR)

CDi = CL**2/(pi*AR)

where
0.11 is the theoretical value of the 2-D lift coefficient per degree of the 2-D section in a frictionless fluid. You might want to use 0.10 instead to compensate for viscosity effects.
Alpha is the angle of attack.
AR is the aspect ratio (= span**2/A).
CDi is the induced drag coefficient.

I have seen these equations in the Larsson book but I am afraid they confuse rather than help me.

To my way of thinking, if lift and drag can be computed by such simple equations then what are the use of programs such as Xfoil, profili or Sailero?

Again it could be that I am missing some point.

In your subsequent mail you have introduced the Moment Coefficient term. This is a term which I have met but I still don't appreciate, yet.

You also quote figures at AoA of 14 degrees. At these values NACA0012 is obviously better. Isn't this value a bit too large for normal sailing conditions?

regards

joe

[sorenfdk]

Originally Posted by farjoe

I have seen these equations in the Larsson book but I am afraid they confuse rather than help me.

To my way of thinking, if lift and drag can be computed by such simple equations then what are the use of programs such as Xfoil, profili or Sailero?

Again it could be that I am missing some point.

In your subsequent mail you have introduced the Moment Coefficient term. This is a term which I have met but I still don't appreciate, yet.

You also quote figures at AoA of 14 degrees. At these values NACA0012 is obviously better. Isn't this value a bit too large for normal sailing conditions?

regards

joe

Once again: Profili/Xfoil are for 2-D calculations only! Keels and rudders are 3-D beasts! You can not use 2-D coefficients to directly calculate the lift and drag of 3-D objects!

The formula for calculating the drag coefficient only gives you the induced drag coefficient (drag that comes from creating lift) - SailAero calculates the total drag including the frictional drag. And it takes stuff like sweep angle and taper ratio into account, something which the two formulas don't.

The moment coefficient is not important here - it was just to show the differences between two ways of doing 2-D calculations. Same goes for the large AoA.

[tspeer]

Originally Posted by farjoe

...To my way of thinking, if lift and drag can be computed by such simple equations then what are the use of programs such as Xfoil, profili or Sailero?...

Ah, the equations may be simple, but where do you get the values of the coefficients to plug into them? That's what aerospace companies and AC campaigns spend millions to get!